You are here


6 posts / 0 new
Last post

dear all,

Does anyone know if RosettaMatch can allow the same residue to satisfy two constraints in RosettaMatch?

thanks for the help!

Post Situation: 
Wed, 2012-01-11 06:50

It can, but only if the sidechain is forming one of the two constraints and the backbone is forming the second of them. If, however, you would like one sidechain to form two contacts, (e.g. OD1 on ASP forming one hbond, and OD2 forming a separate hbond) then, I believe you'd need to define that as a single constraint with a certain restricted set of acceptable geometries and run a separate matching job with that alternate constraint.

Sun, 2012-01-15 15:09

thanks for the reply. That makes sense as one constraint would fully define the interaction anyway I suppose.

Mon, 2012-01-16 02:35

Sorry, I have another question!

When I use the -use_input_sc flag in the match app I assume this means that it will include the pre-existing sidechain rotamers in the search. I would have thought that when searching a scaffold using constraints derived from the same scaffold, I would always be able to retrieve the correct real active site when using the -use_input_sc flag.

However, this does not seem to be the case at the moment. I have 6 constraints and I am failing to find any solutions.

Of course I could have other problems with my constraints but does the -use_input_sc flag actually do what I think it does? Or does it have no effect in the match app?

Mon, 2012-01-16 10:01

I wrote a separate application at one point, but that isn't in the release, to measure the exact geometries of the native contacts. There were a few examples I stumbled upon where the native rotamers with the "idealized" contact geometries failed to recapitulate the native sequence. For your problem, assuming you want to embed your contacts in a new scaffold, I would suggest ramping up the amount of sampling you're doing. To recover the native, I would make sure your geometric constraint file contains the native geometry.

Here's the app to measure the native geometry: really simple and stupid, but useful. I haven't tried to compile it for a while, so it may require a bit of tweaking, but it should get you 99% of the way there. (Rename it -- I can only upload .txt files.)


Tue, 2012-01-17 07:13

I had written a similar program myself in fact :)
I was just searching a known enzyme using exact native geometries taken from the same enzyme in order to see if I could recapitulate the correct active site positions. This was only for curiosity rather than to do anything useful because I was having problems finding hits in other scaffolds. I found this works for subsets of my constraints but not if I include all constraints. I assumed this was due to rotamer sampling so used the flags -ex1 -ex2 -ex3 etc, but I was wondering if the -use_input_sc flag is ignored in the match app?

Anyway thanks for the help!

Thu, 2012-01-19 02:49